Recently, a new model for the evolution of female mating preferences has been proposed and tested. This model hypothesizes that biases favoring a male trait can be established in sensory and/or cognitive systems prior to the appearance of the male trait. These "preexisting biases" can subsequently result in selection on male traits when they arise. Using the comparative method, it has been shown that a female preference favoring a male trait, a sword, arose prior to the evolution of the sword in Xiphophorus fishes: a number of species of Xiphophorus have both a preference favoring a sword and a sword, while the sister genus, Priapella has a sword preference, but lacks a sword. Phylogenetic analysis suggests then that a preference favoring a sword arose prior to the sword's appearance. The sword is a complex trait; a female preference for such a complex trait may have more than one proximate basis. For example, a preference may be for specific traits, may be a general preference for conspicuousness, or may have many bases. The long-term goal of this research is to examine in detail the preexisting bias favoring the sword, and compare the bias favoring the sword among three groups: species lacking swords, species which possess 1-3 components of the sword, and species with complete swords. This will allow me to determine not only the structure of the preference, but also if and how the bias has been modified through evolutionary history (it is possible that the sword bias in sworded species is stronger than in unsworded species due to coevolutionary processes after the origin of the sword components). The specific aim of the one-year B/START project proposed here will be to examine in detail several hypotheses designed to elucidate the base(s) of the preexisting bias for a sword-in a species lacking a sword. To do this, I will use digital video technology to manipulate the sword components and several types of caudal fin elaboration; by examining the efficacy of different types of stimuli in eliciting mating responses from females, I will develop a detailed characterization of the preexisting bias favoring a sword. The proposed study is important because it will investigate the proximate basis of a female preference. While numerous studies have shown that females prefer some trait variants over others, the nature of the preference is almost never known. The results from these experiments will contribute to our understanding of the nature of a preexisting bias prior to the evolution of traits it selects. The research proposed here addresses one of the National Institute of Mental Health areas of interest in behavioral neuroscience research outlined in the Comparative Approaches to Brain and Behavior Program Announcement; this project will identify significant stimuli that elicit patterns of behavior. In addition, the digital technology proposed in this project greatly reduces the number of test subjects necessary to obtain significant results, as well as circumvents the need of invasive manipulations of phenotypic traits.